Enhancement of the Electron Field Emission Properties of

Aug 5, 2014 - Growth of graphene on Cu foils by microwave plasma chemical vapor ... A tunable microplasma gradient-index lens for millimeter waves...
1 downloads 0 Views 4MB Size
Download PDF
Research Article www.acsami.org

Enhancement of the Electron Field Emission Properties of Ultrananocrystalline Diamond Films via Hydrogen Post-Treatment Kamatchi Jothiramalingam Sankaran,† Srinivasu Kunuku,‡ Keh-Chyang Leou,‡ Nyan-Hwa Tai,*,† and I-Nan Lin*,§ †

Department of Materials Science and Engineering and ‡Department of Engineering and System Science, National Tsing Hua University, Hsinchu 300, Taiwan, Republic of China § Department of Physics, Tamkang University, Tamsui 251, Taiwan, Republic of China ABSTRACT: Enhanced electron field emission (EFE) properties due to hydrogen post-treatment at 600 °C have been observed for ultrananocrystalline diamond (UNCD) films. The EFE properties of H2-gas-treated UNCD films could be turned on at a low field of 5.3 V/μm, obtaining an EFE current density of 3.6 mA/cm2 at an applied field of 11.7 V/μm that is superior to those of UNCD films treated with H2 plasma. Transmission electron microscopic investigations revealed that H2 plasma treatment induced amorphous carbon (a-C) (and graphitic) phases only on the surface region of the UNCD films but the interior region of the UNCD films still contained very small amounts of a-C (and graphitic) grain boundary phases, resulting in a resistive transport path and inferior EFE properties. On the other hand, H2 gas treatment induces a-C (and graphitic) phases along the grain boundary throughout the thickness of the UNCD films, resulting in creation of conduction channels for the electrons to transport from the bottom of the films to the top and hence the superior EFE properties. KEYWORDS: ultrananocrystalline diamond films, hydrogen post-treatments, electron field emission, plasma illumination, electron energy loss spectroscopy

■

phase grain boundaries10,11 would form possible interconnected paths for facilitating the transport of electrons and markedly enhance the EFE properties. In addition, the modification of surface characteristics of diamond films plays an important role in enhancing the EFE behavior of the films. The EFE properties of diamond films were markedly altered when coated with wide-band-gap materials.12 Even when the surface of a diamond film is bonded with different species, the band structure below the surface is altered, which changes the EFE properties. The type of surface treatment influences the surface conductivity and electron affinity properties. For example, hydrogen treatment of diamond shows p-type surface conductivity with NEA,13,14 whereas oxygen treatment of diamond induced surfaceinsulating properties with a positive electron affinity.15,16 The change of surface treatment of diamond (O2 or H2) is reversible and controllable, which has created a niche application for diamond in sensors, biology, and electronics applications.17−19 The hydrogen treatment is of particular interesting owing to its simplicity and versatility. Hydrogen treatment of diamond usually has been achieved using atomic hydrogen produced by either plasma or hot filament techniques20,21 and was lately also

INTRODUCTION Diamond and related materials have enormous prospective applications due to their excellent physical and chemical properties.1,2 Among diamond films with various granular structures, the ultrananocrystalline diamond (UNCD) film is a more promising material due to its unique and advantageous properties similar to those of conventional microcrystalline and nanocrystalline diamond films, such as high hardness and smoothness, chemical inertness, negative electron affinity (NEA) characteristics, high transmittance in the near-infrared region, and the likelihood of p- and n-type conductivities by doping.3−6 The UNCD films are superior to conventional diamond films in many ways, which widespread research has confirmed. Moreover, nanosizing of diamond crystals increases the grain boundaries, containing nondiamond carbons, in the UNCD films and has been observed to markedly enhance the electron field emission (EFE) properties of the films.7,8 UNCD films with ultrasmall grain sizes of 5−10 nm and sp2-rich grain boundaries show better EFE characteristics as compared to the conventional diamond films.9 Consequently, UNCD films possess better potential application for cold cathode field emitters and vacuum nanoelectronic devices. Appropriate modification and control of microstructures of UNCD films can modify their physical and chemical properties for emerging new applications and technologies. Prior studies have reported that a proper increase in amorphous carbon (a-C) and graphitic © 2014 American Chemical Society

Received: June 16, 2014 Accepted: August 5, 2014 Published: August 5, 2014 14543

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551

ACS Applied Materials & Interfaces

Research Article

done electrochemically.22 Manfredotti et al. observed that the thermal hydrogenation can give rise to a surface conducting layer with the same electrical characteristics observed in plasma hydrogenation.23 Ando et al. demonstrated the possibility of thermal hydrogen by exposing diamond powders to a pure molecular hydrogen flow in the temperature range of 400− 1100 °C.24 Fizzotti et al. reported that thermal hydrogenation of diamond using molecular hydrogen on single-crystal diamond at 800 °C gave rise to surface conduction and was interpreted through the electrochemical transfer-doping model.25 Particularly, enhancement of the EFE properties was observed for the surface-treated UNCD films.26−28 However, these reports on the effect of H2 post-treatment of the electrical conductivity/EFE properties of the diamond films are controversial, and the mechanism by which the H2 posttreatment process influenced the EFE properties of these films has not been fully understood yet. In this study, the UNCD films were subjected to two different hydrogen post-treatments: (1) hydrogen plasma treatment of the UNCD films was performed in a microwave plasma enhanced chemical vapor deposition (MPECVD) system, and (2) hydrogen gas treatment was carried out by placing the UNCD films in a nonplasma quartz tube reactor connected with a H2 gas supply line at low treatment temperature. We observed a significant effect of these two post-treatment processes on enhancing the plasma illumination (PI) and the EFE properties of the UNCD films. The modifications to the bonding structure of these films due to these hydrogen post-treatment processes were investigated in detail by electron energy loss spectroscopy (EELS) in transmission electron microscopy (TEM). A possible mechanism for the enhancement of EFE and PI properties of UNCD films is discussed.

■

a four-probe technique, in which two configurations were utilized: (i) with all four probes located on the surface of the films for measuring the resistivity along the film surface and (ii) with two of the probes located on the UNCD film surface and the other two probes located on the UNCD-to-Si interface for measuring the resistivity through the film thickness. The EFE properties of the UNCD films were measured with a tunable parallel-plate setup, in which the cathode-to-anode distance was controlled using a micrometer. The current density− applied field (Je−E) characteristics were measured using an electrometer (Keithley 2410) under a pressure below 10−6 Torr (1.3 × 10−4 Pa). The EFE behavior of the materials was modeled with the Fowler− Nordheim (F−N) theory.29 The turn-on field (E0) was designated as the point of interception of the straight lines extrapolated from the low- and high-field segments of the F−N plots, namely, log(Je/E2) versus 1/E plots. The PI characteristics of a microcavity were also investigated, in which an indium tin oxide (ITO)-coated glass was used as the anode and the UNCD films were used as the cathode. The cathode-to-anode separation was fixed by a poly(tetrafluoroethylene) (PTFE) spacer (1.0 mm in thickness). A circular hole about 6.0 mm in diameter was cut from the spacer (PTFE) to shape a microcavity. The chamber was evacuated to reach a base pressure of 0.1 mTorr (13 mPa) and then purged with Ar for 10 min. The Ar gas was flowed into the chamber at a flow rate of 10 sccm throughout the measurements. The plasma was triggered using a pulsed direct current voltage in a bipolar pulse mode (20 ms square pulse, 6 kHz repetition rate) at 12 Torr (1599 Pa). The plasma current versus applied voltage behavior was measured using an electrometer (Keithley 237).

■

RESULTS AND DISCUSSION Figure 1 shows the apparent changes in the morphology of the films due to hydrogen treatments. The pristine UNCD films

EXPERIMENTAL METHODS

The UNCD films were grown on n-type mirror polished Si substrates in an MPECVD (2.45 GHz 6 in. IPLAS-CYRANNUS-I, Troisdorf, Germany) system. Prior to the deposition of UNCD films, the silicon substrates were preseeded by ultrasonication in methanol solution, containing nanosized diamond powders (∼5 nm) and Ti powders (∼32.5 nm) for 45 min. The substrates were ultrasonicated again in methanol for 1 min to remove the possibly adhered nanoparticles. The UNCD films were grown in Ar (99%)/CH4 (1%) plasma (100 sccm) at a microwave power of 1200 W and at a pressure of 120 Torr (16 kPa) for 5 h, without the substrate being heated. The substrate was heated due to the bombardment of the plasma species, and the substrate temperature was estimated to be around 475 °C during the growth of the UNCD films. The pristine UNCD films were subjected to two different hydrogen post-treatment processes. In the first method, the pristine UNCD films were H2-plasma-treated in the same IPLAS-CYRANNUS-I MPECVD system for 5 min at a microwave power of 600 W, a pressure of 5 Torr (666 Pa) (with 100 sccm of H2 flow), and a temperature of 600 °C. Thus, the obtained samples are designated as “H2-plasma-treated UNCD” films. In the other process, the UNCD films were thermally heat-treated in a H2 environment; i.e., the UNCD films were heated in a nonplasma quartz tube reactor to 600 °C under H2 gas (99.999% pure flow, 20 sccm) for 2 h, followed by cooling to room temperature under a continuous H2 gas flow. Thus, the obtained samples are designated as “H2-gas-treated UNCD” films. The films were characterized using field emission scanning electron microscopy (FESEM; Jeol 6500), atomic force microscopy (AFM; Smena-B, NT-MDT, Russia), and visible-Raman spectroscopy (λ = 632.8 nm, spectral resolution 0.5 cm−1, Lab Raman HR800, Jobin Yvon). The detailed microstructure and bonding structure of the samples were examined using EELS (Gatan Enfina) in TEM (Jeol 2100F). The electrical resistivity of the UNCD films was measured by

Figure 1. FESEM images of (a) pristine, (b) H2-plasma-treated, and (c) H2-gas-treated UNCD films. The insets show the corresponding cross-sectional FESEM images.

contain random and spherically structured grains (Figure 1a). The thickness of the pristine UNCD was about 400 nm, which was estimated from a cross-sectional FESEM image (inset, Figure 1a). The root-mean-square (rms) roughness of the pristine UNCD films was 7.8 nm, which was estimated using AFM (figure not shown). More distinct and clear grains and 14544

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551

ACS Applied Materials & Interfaces

Research Article

structure of these films using TEM/EELS is necessary, which will be discussed shortly. The electrical resistivity of the UNCD films was measured by the four-probe technique. Figure 3a indicates the resistivity

grain boundaries were observed after 5 min of H2 plasma treatment due to etching of nondiamond carbon from the surface of the UNCD films (Figure 1b). The thickness (inset, Figure 1b) and the rms roughness (figure not shown) of the H2-plasma-treated UNCD films were reduced to 350 and 6.7 nm. Agglomeration of the nanosized grains was induced for the UNCD films post-treated in H2 gas, resulting in the elongated granular morphology (Figure 1c). The thickness of the films was reduced to around 380 nm (inset, Figure 1c), and the rms roughness was estimated to be 8.9 nm. It seems that both the H2 plasma and H2 gas post-treatment processes resulted in some etching on the surface of the diamond films. Figure 2 depicts the visible−Raman spectra of (I) pristine, (II) H2-plasma-treated, and (III) H2-gas-treated UNCD films.

Figure 2. Visible−Raman spectra from (I) pristine, (II) H2-plasmatreated, and (III) H2-gas-treated UNCD films taken at 632.8 nm.

Figure 3. Electrical resistivity (current−voltage curves) measured by the four-probe technique for (I) pristine, (II) H2-plasma-treated, and (III) H2-gas-treated UNCD films. The current was flowing either (a) along the surface of the films (i.e., with all four probes on the same surface of the films, inset) or (b) across the films (i.e., with two of the probes on the surface of the diamond films and two on the Si substrates, inset).

The Raman spectra were deconvoluted by using teh multipeak Lorentzian fitting method. Peaks at around 1160 and 1470 cm−1 are attributed to the ν1 and ν3 modes of transpolyacetylene (t-PA) present in the grain boundaries of pristine UNCD films (curve I, Figure 2).30 A broadened peak around 1340 cm−1 (designated as the D* band), which correlates to the disordered sp2-bonded carbon, is observed for the films.31 The sharp Raman peak at 1332 cm−1 corresponding to the F2g zone center optical phonon of diamond is not clearly observable in this Raman spectrum as visible−Raman spectroscopy is overwhelmingly more sensitive to sp2 sites. The G band of the UNCD film is observed at around 1540 cm−1. A shoulder peak around 1607 cm−1 (designated as the D′ band) is seen, which possibly arises from the nanocrystalline graphitic content in the films.32 Curves II and III of Figure 2 show the effect of H2 post-treatments on altering the bonding characteristics of the UNCD films. They show higher intensities of the D* and G peaks, indicating amorphization and graphitization types of transitions.33 Particularly, the ν3 peak of H2-gas-treated UNCD films is not clearly observed as it merged with the G band due to the use of a red laser, as can be seen in curve III of Figure 2. Interestingly, there is a big shift in the G and ν3 peaks for H2gas-treated UNCD films, indicating that the increase in a-C (or graphite) phases of the H2-gas-treated UNCD films is higher as compared to that of the H2-plasma-treated UNCD films.34 However, the change in Raman structure cannot unambiguously differentiate the effect of H2 plasma and H2 gas treatment processes on altering the characteristics of UNCD films. A more detailed investigation on the microstructure/bonding

along the UNCD film surface (ρs), whereas Figure 3b shows the resistivity through the film thickness (ρt). The configurations for measuring ρs and ρt are shown in the respective insets. This figure indicates that both the H2-post-treated UNCD films show markedly better conductivity compared to the pristine UNCD films. Figure 3a reveals that the surface resistivity (ρs) is very large (0.15 Ω cm, curve I of Figure 3a) for pristine UNCD films. The H2 plasma and H2 gas treatment processes markedly lowered the surface resistivity of the UNCD films. The ρs value of the H2-plasma-treated UNCD (0.022 Ω cm) is similar to those of the H2-gas-treated UNCD films (0.024 Ω cm) (curves II and III, Figure 3a). In contrast, Figure 3b shows the through-film resistivity (ρt) is around 0.46 × 10−4 Ω cm for the H2-gas-treated UNCD films (curve III, Figure 3b), which is noticeably smaller than the ρt value for H2plasma-treated UNCD films (1.15 × 10−3 Ω cm, curve II of Figure 3b). The pristine UNCD films show markedly larger through-film resistivity (2.68 × 10−2 Ω cm, curve I of Figure 3b). These resistivity values with the percentage of error in measurements are listed in Table 1. Figure 4a shows the EFE properties of these UNCD films, with the inset showing the corresponding F−N plots. As seen from the Je versus E curves, the EFE properties are significantly enhanced due to the H2 post-treatment processes (curves II 14545

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551

ACS Applied Materials & Interfaces

Research Article

Table 1. Electrical Resistivity (Current−Voltage Curves) Measured by the Four-Probe Technique, Electron Field Emission, and Plasma Illumination Properties of UNCD Films That Have Undergone Different Hydrogen Treatmentsa electron field emission

four-probe measurements ρt (Ω cm) (error, %)

ρs (Ω cm) (error, %)

E0 (V/μm) (error, %)

pristine

2.68 × 10−2 (3.5)

0.15 (7.5)

18.5 (7.8)

H2-plasma-treated

1.15 × 10−3 (9.1)

0.022 (8.9)

10.4 (8.9)

H2-gas-treated

0.46 × 10−4 (6.2)

0.024 (4.9)

5.3 (4.9)

UNCD film

Je (mA/cm2) (error, %) 1.4 (8.0) at 35.7 V/μm 2.3 (9.2) at 21.8 V/μm 3.6 (7.5) at 11.7 V/μm

plasma illumination Eth (V/μm) (error, %)

Jpl (mA/cm2) at 0.55 V/μm (error, %)

ne (/cm3) at 5000 V/cm (error, %)

0.35 (4.5)

0.47 (5.6)

1.87 × 1016 (5.8)

0.33 (6.3)

0.50 (6.2)

2.10 × 1016 (6.2)

0.31 (7.5)

0.54 (6.8)

2.46 × 1016 (7.1)

ρt is the through-film resistivity measured using two probes placed on the UNCD surface and the other two probes on the UNCD-to-Si interface. ρs is the surface resistivity measured using all four probes placed on the UNCD surface. E0 is the turn-on field for the EFE process that was designated as the interception of the lines extrapolated from the high-field and low-field segments of the F−N plots. Je is the EFE current density evaluated at the applied field designated. Eth is the threshold field needed to trigger the plasma for the microplasma devices. Jpl is the plasma current density at 0.55 V/μm applied field for the microplasma devices. ne is the plasma density estimated from the plasma current density at 5000 V/cm applied field for the microplasma devices. a

and III, Figure 4a), as compared to those of pristine UNCD films (curve I, Figure 4a). Better EFE properties are observed for H2-gas-treated UNCD films compared to H2-plasma-treated UNCD films. The E0 value was lowered from 18.5 V/μm for pristine UNCD films to 5.3 V/μm for H2-gas-treated UNCD films. A better Je value of 3.6 mA/cm2 at an applied field (Ea) of 11.7 V/μm is obtained for H2-gas-treated UNCD films (curve III, Figure 4a), whereas the Je value for pristine UNCD films is markedly smaller, i.e., Je = 1.4 mA/cm2, at Ea = 35.7 V/μm. These vaules are given in Table 1, including the percentage of error in the measurements. In contrast, for H2-plasma-treated UNCD films (curve II, Figure 4a), the E0 value is 10.4 V/μm and the Je value is 2.3 mA/cm2 at Ea = 21.8 V/μm, and these values are markedly better than those for the pristine UNCD films but are still inferior to those for the H2-gas-treated UNCD films. The comparison of the EFE properties of H2-gas-treated UNCD films with those of other plasma-treated electron field emitting materials is summarized in Table 2.35−38 The Table 2. Comparison of Electron Field Emission Properties of H2-Gas-Treated UNCD Films with Those of Various Plasma-Treated Electron Field Emitting Materials electron field emitting material 28

H2-plama-treated diamond nanowires H2-plasma-treated tungsten oxide nanowires35 H2-plasma-treated ZnO films36 Ar-plasma-treated ultrasharp α-Fe2O3 nanoflakes37 H2-plasma-treated diamond-like carbon38 H2-gas-treated UNCD films (present study)

E0 (V/μm)

Je (mA/cm2)

4.2 4.7

5.1 at 8.5 V/μm 0.15 at 6.0 V/μm

3.6 8.0

16.0 at 1.3 V/μm 0.06 at 11.0 V/μm

1000 5.3

3.6 at 11.7 V/μm

enhancement in EFE properties is closely related to the improvement in the electrical conductivity of the films. Restated, both the better surface and bulk electrical characteristics of the H2-gas-treated UNCD films distinctly improved the EFE properties. On the other hand, Figure 5 shows a series of photographs of the PI behavior of the microplasma devices, which utilized the UNCD films as cathode materials. The intensity of the plasma illumination increases monotonically with the applied voltage. These figures show that the devices using H2-gas-treated UNCD films as cathodes (Figure 5c) exhibit superior illumination behavior compared to those using pristine and

Figure 4. (a) Electron field emission (with corresponding F−N plots shown as an inset), (b) plasma current density against applied field curves, and (c) plasma density against applied field curves of (I) pristine, (II) H2-plasma-treated, and (III) H2-gas-treated UNCD films.

14546

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551

ACS Applied Materials & Interfaces

Research Article

ne =

H2-plasma-treated UNCD films as cathodes (Figure 5b) and both of the devices using treated films as cathodes show better PI behavior than those using the pristine films as cathodes (Figure 5a). The PI characteristics of the plasma devices are better illustrated by the variation of the plasma current density (Jpl) versus E, which is plotted in Figure 4b. The microplasma devices based on the H2-gas-treated UNCD films need 310 V (threshold field of (Eth) 0.31 V/μm, Figure 5c) to ignite the plasma, reaching a Jpl value of 0.54 mA/cm2 at Ea = 0.55 V/μm (curve III, Figure 4b). In contrast, the microplasma devices based on H2-plasma-treated UNCD films can be triggered at 330 V (Eth = 0.33 V/μm, Figure 5b) with Jpl = 0.50 mA/cm2 (at Ea = 0.55 V/μm) (curve II, Figure 4b), whereas those based on pristine UNCD films required larger voltage of 350 V (Eth = 0.35 V/μm, Figure 5a) to trigger the plasma, with Jpl = 0.47 mA/cm2 at Ea = 0.55 V/μm (curve I, Figure 4b). The values of Eth and Jpl are extracted from these figures and are also listed in Table 1, including the percentage of error in the measurements. The trend is similar to the change in EFE behavior (cf. Figure 4a). Restated, the best plasma illumination behavior, i.e., the lowest Eth value of 0.31 V/μm and the largest Jpl value of 0.54 mA/cm2 at Ea = 0.55 V/μm, are observed for the microplasma devices based on H2-gas-treated UNCD films. In the meantime, we have estimated the plasma density (ne) by using Child’s law, shown in the following equation:39

shows that the surface region of the pristine UNCD films consists of ultrasmall spherical diamond grains (5−10 nm). The diffraction rings in the selected area electron diffraction (SAED) pattern (inset of Figure 6aI) are smooth and continuous, indicating that the diamond grains are randomly oriented. The BF image shown in Figure 6bI indicates that, for the surface region of H2-plasma-treated UNCD films, large diamond aggregates about hundreds of nanometer in size were induced and were evenly distributed among the matrix of ultrasmall grains, whereas Figure 6cI reveals that, for the surface region of H2-gas-treated UNCD films, only a mild agglomeration process was induced, resulting in elongation of the diamond grains, which are around tens of nanometers in size. The SAED patterns of H2-plasma- and H2-gas-treated UNCD

(1)

Here ne is the electron density, Jc is the plasma current density at the cathode, e is the electron charge, and uB is the Bhom velocity. The Bhom velocity of Ar ion strictly depends on the kinetic energy of the electron and the mass of the ion, which is described by the following equation:40

uB =

kTe mi

(3)

Figure 6. (aI, bI, cI) Bright field and (aII, bII, cII) composed dark field TEM micrographs of the surface regions of UNCD films (with the insets showing the SAED patterns). The samples were (aI, aII) pristine, (bI, bII) H2-plasma-treated, and (cI, cII) H2-gas-treated UNCD films.

Jc euB

e kTe

By using the above formulas, we have obtained ne versus E for H2-gas-treated UNCD films (curve III, Figure 4c), achieving a value of (ne)III = 2.46 × 1016/m3 at V = 500 V and a pressure of 2 Torr (266 Pa). Similarly, we have obtained an ne value of 1.87 × 1016/m3 at V = 500 V for pristine diamond films (curve I, Figure 4c) and (ne)II = 2.10 × 1016/m3 at V = 500 V for H2plasma-treated UNCD films (curve II, Figure 4c). These values are also listed in Table 1. While the SEM micrographs and Raman spectra provide valuable information on the evolution of the surface morphology and bonding structure of the UNCD films due to H2 post-treatment processes, how these processes modified the electric conductivity/EFE properties of the UNCD films is still not explained. This requires an investigation on the evolution of the UNCD’s granular structure to understand the related mechanism. For this purpose, the samples were investigated using TEM and EELS (in TEM). It should be noted that when the samples were ion-milled only from the Si side, the TEM foil contained mostly the regions near the surface of the UNCD films, whereas when the samples were ion-milled both from the top and the bottom surfaces evenly at the same time, the thin foil contained the materials in the interior region of the UNCD films. The bright field (BF) and the associated dark field (DF) TEM micrographs for the surface region of the UNCD films are shown in Figure 6. Figure 6aI

Figure 5. Plasma illumination characteristics for the microplasma devices utilizing (a) pristine, (b) H2-plasma-treated, and (c) H2-gastreated UNCD films as the cathode.

ne =

Jc m i

(2)

Here k is the Boltzmann constant and Te is the electron temperature (7841 K).41 By substituting the Bohm velocity in eq 1, we can obtain the complete equation for estimation of the electron density: 14547

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551

ACS Applied Materials & Interfaces

Research Article

films (insets of parts bI and cI, respectively, of Figure 6) contain spotty diffraction rings, implying again that these films contain some large diamond aggregates coexisting with the ultrasmall diamond grains. The size distribution of the diamond grains in these films is better illustrated by the composed dark field (cDF) images, which are the superposition of several dark field images acquired using different diffraction spots (indicated in the inset SAED pattern). Parts aII, bII, and cII of Figure 6 show the c-DF images of pristine, H2-plasma-treated, and H2-gastreated UNCD films, corresponding to the BF images in parts aI, bI, and cI, respectively, of Figure 6. These c-DF images indicate again that the H2 plasma treatment process induced more marked coalescence of the ultrasmall diamond grains compared with the H2 gas treatment process. Figure 7 shows the TEM micrographs of the interior regions of the UNCD films, indicating that the interior region of all the

simultaneous attack of the diamond grains and the grain boundaries will smooth out the surface roughness of the films. In contrast, the H2 gas does not attack the diamond grains at all. Thus, the UNCD films were thermally heat-treated in a H2 environment (1 atm) at 600 °C for 2 h for “H2 gas” films. However, the H2 gas can diffuse in and out of the UNCD films easily43 and will interact with the grain boundary phase throughout the whole thickness of the UCND films. The preferential etching on the grain boundaries results in an increase in the roughness of the films. The different behaviors in modifying the microstructure of the UNCD films due to H2 plasma and H2 gas post-treatment processes is interesting, but such information is still not able to account for the different degrees of improvement in the electrical conductivity/EFE properties of the UNCD films due to these processes. To understand the genuine mechanism of the enhancement of those properties, the bonding structure of the UNCD films was examined using EELS. Figure 8 shows the

Figure 7. (aI, bI, cI) Bright field and (aII, bII, cII) composed dark field TEM micrographs of the interior regions of UNCD films (with the insets showing the SAED patterns). The samples were (aI, aII) pristine, (bI, bII) H2-plasma-treated, and (cI, cII) H2-gas-treated UNCD films.

films shows a granular structure similar to that observed in the surface region of the films (cf. Figure 6); i.e., there are large diamond aggregates distributed evenly among the matrix of nanosized diamond grains. However, more detailed examinations reveal a subtle difference in the microstructure of the interior region due to different H2 post-treatment processes. While the interior region of the H2-gas-treated UNCD films (Figure 7cI) is very similar to the surface region, the interior region of the H2-plasma-treated UNCD films (Figure 7bI) contains smaller aggregates compared to the surface region. Again, the c-DF images were acquired to more clearly show the difference in granular structure of the interior regions of the UNCD films compared to the surface regions of the films as illustrated in Figure 7aII,bII,cII, corresponding to the BF images in Figure 7aI,bI,cI, respectively. It should be pointed out that the H2 plasma is much more aggressive in attacking the UNCD films as compared with the H2 gas, since the H2 plasma contains a large percentage of atomic hydrogen. However, due to the presence of sheath in the plasma, the plasma cannot penetrate deep into the diamond films; i.e., the interaction zone between the H2 plasma and the UNCD films is very shallow. The H2 plasma can not only attack the grain boundary phase but also etch the diamond grains.42 Therefore, for “H2 plasma” films, the UNCD films were treated in H2 plasma for only 5 min, as a too long H2 plasma treatment process might completely etch away the UNCD films. The

Figure 8. Selected area EELS spectra of the surface regions of UNCD films: (a) core-loss EELS spectra and (b) plasmon-loss EELS spectra of (I) pristine, (II) H2-plasma-treated, and (III) H2-gas-treated UNCD films corresponding to the TEM micrographs of Figure 6.

selected area EELS spectra corresponding to each micrograph in Figure 6, revealing that, in the surface region of the UNCD films, there are significant changes in the bonding structure due to the post-treatment processes. The carbon edge core-loss EELS spectra corresponding to these UNCD films (Figure 8a) indicate the presence of an abrupt rise near 289.5 eV (σ* band) and a large dip in the vicinity of 302 eV, implying that the diamond nature of these materials was not altered due to the post-treatment processes.44,45 Moreover, there is a prominent π* band at 285.5 eV in core-loss EELS spectra for H2-gastreated (curve III, Figure 8a) and H2-plasma-treated (curve II, Figure 8a) UNCD films, but this band is not visible in the spectra for pristine UNCD films (curve I, Figure 8a), indicating that a large proportion of sp2-bonded carbon was induced, probably along the grain boundary regions of the UNCD films due to the H2 gas and H2 plasma post-treatment processes. However, it is still necessary to differentiate the nature of the 14548

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551

ACS Applied Materials & Interfaces

Research Article

sp2-bonded carbon to understand the genuine mechanism for enhancing the EFE/PI properties of the UNCD films due to these H2 post-treatment processes. It should be pointed out that plasmon-loss EELS is the most effective measurement for distinguishing the crystalline sp2-bonded carbons (the graphite) from the amorphous ones, as the plasmon-loss EELS spectrum for the graphitic phase shows a prominent peak at s3 (27 eV) and that for the a-C phase shows a peak at s1 (22 eV).45,46 In contrast, the crystalline sp3-bonded carbons, the diamond, show a peak due to the bulk plasmon at s4 (33 eV) with a shoulder corresponding to the surface plasmon at s2 (23 eV). The Is2/Is4 ratio is about 1/√2. Figure 8b shows the plasmon-loss EELS spectra of the same regions. It is observed that the pristine UNCD films (curve I, Figure 8b) are dominated by s2 and s4 peaks, indicating that pristine UNCD films are diamond.45 In contrast, the plasmonloss EELS spectrum of H2-plasma treated UNCD films (curve II, Figure 8b) is dominated by a larger s3 band (∼27 eV) with s2 and s4 bands of much smaller intensity, indicating that this region consists of some proportion of graphitic phases besides diamond. Quite the opposite, the plasmon-loss spectrum for H2-gas-treated UNCD films (curve III, Figure 8b) is subjugated by an s1 band (22 eV), in which s2 and s4 bands still dominate but with increased spectral weight of the s1 band (22 eV). This result indicates that H2 gas treatment induces a-C phases along with some graphitic phases in the UNCD films. From the fourprobe measurements, it is observed that H2 gas and H2 plasma treatment processes resulted in similar low surface resistivity values (cf. Figure 3a) compared with those of the pristine UNCD films. Such a four-probe observation is in accord with the phenomena that both H2 gas and H2 plasma posttreatments induced the formation of a-C (and graphitic) phases on the surface region of the UNCD films. In contrast, the four-probe measurements showed that H2gas-treated UNCD films possess a markedly lower through-film resistivity value than the H2-plasma-treated UNCD films (cf. Figure 3b). The question now is how the H2-gas-treated UNCD films show better through-film electrical conductivity/ EFE properties than the H2-plasma treated films. Apparently, an EELS analysis on the interior regions of the UNCD films is required to understand how the through-film conductivity was enhanced due to these H2 post-treatment processes. The carbon edge core-loss EELS spectra (Figure 9a), corresponding to each TEM micrograph of Figure 7, indicate the presence of an abrupt rise near 290.5 eV (σ* band) and a large dip in the vicinity of 302 eV for all the UNCD films, revealing that the diamond nature of the UNCD films was not altered for the interior region. However, it is interesting to observe that the prominent π* band at 285.5 eV in the interior-loss EELS spectrum is observed only for H2-gas-treated UNCD films (curve III, Figure 9a) and is not visible in the spectrum for either the pristine or the H2-plasma-treated UNCD films (curves I and II, respectively, Figure 9a). The plasmon-loss EELS spectra of the interior region of the UNCD films are shown in Figure 9b. The spectrum of the interior region of the pristine UNCD films (curve I, Figure 9b) shows behavior similar to that of the spectrum of the surface region in which s2 and s4 peaks are the dominant peaks, indicating that the pristine UNCD films are diamond.45 There is no sp2-bonded carbon in this region. Surprisingly, the plasmon-loss EELS spectrum of the interior region of the H2plasma-treated UNCD films (curve II, Figure 9b) also indicates

Figure 9. Selected area EELS spectra of the interior regions of UNCD films: (a) core-loss EELS spectra and (b) plasmon-loss EELS spectra of (I) pristine, (II) H2-plasma-treated, and (III) H2-gas-treated UNCD films corresponding to the TEM micrograph of Figure 7.

that s2 and s4 peaks are the dominant peaks, as for pristine UNCD films, with no presence of the s3 band (∼27 eV) of graphitic phases or s1 band (∼22 eV) of a-C phases, which is totally different from the case for the surface region of H2plasma-treated UNCD films (cf. curve II, Figure 8b). Here it is evident that H2 plasma treatment induces the graphitic phases only in the surface region and not in the interior region of the UNCD films. In contrast, the plasmon-loss spectrum of the interior region of the H2-gas-treated UNCD films (curve III, Figure 9b) shows that this region is still dominated by an s1 band (22 eV) (including s2 and s4 bands of less intensity). Restated, the TEM−EELS analyses reveal that H2 gas treatment induces a-C (and graphite) phases throughout the whole UNCD films, whereas the H2 plasma treatment induces graphitic phases only at the surface region of the UNCD films. Previous reports revealed that the a-C (or graphitic) phases are more conducting than the t-PA phases,47,48 such that the formation of sp2-bonded carbon phases at the grain boundaries creates conduction channels for the electrons to transport from the bottom side to the top side and emit to vacuum very easily. In this work, H2 gas treatment induces a-C (or graphite) phases throughout the whole region of the UNCD films such that the electrons can be transported easily along the a-C (or graphite) phases at the grain boundaries throughout the diamond films to the emitting surface and are then emitted to vacuum without any difficulty as the diamond surfaces are NEA in nature.49−51 On the contrary, for the H2-plasma-treated UNCD films, the aC (or graphitic) phases are present only at the surface region, and the interior region of these films remained essentially as diamond grains with lower amounts of a-C (or graphite) phases in the grain boundaries. This is the reason that the electrons cannot transport easily from the bottom side of the sample to the top side, resulting in inferior EFE properties. Consequently, the formation of a-C (or graphite) phases along the grain boundaries throughout the whole UNCD films due to H2 gas treatment is the genuine reason for the enhanced EFE and PI 14549

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551

ACS Applied Materials & Interfaces

Research Article

properties of these UNCD films. The formation of point defects seems not to be the source of enhancement of the EFE properties of H2-treated UNCD films.

Electron Field Emission of Ultrananocrystalline Diamond Films. J. Appl. Phys. 2001, 89, 2958−2967. (5) Sharda, T.; Rahaman, M. M.; Nukaya, Y.; Soga, T.; Jimbo, T.; Umeno, M. Structural and Optical Properties of Diamond and NanoDiamond Films Grown by Microwave Plasma Chemical Vapor Deposition. Diamond Relat. Mater. 2001, 10, 561−567. (6) Williams, O. A.; Curat, S.; Gerbi, J. E.; Gruen, D. M.; Jackman, R. B. n-Type Conductivity in Ultrananocrystalline Diamond Films. Appl. Phys. Lett. 2004, 85, 1680−1682. (7) Sankaran, K. J.; Joseph, P. T.; Chen, H. C.; Tai, N. H.; Lin, I. N. Investigation in the Role of Hydrogen on the Properties of Diamond Films Grown Using Ar/H2/CH4 Microwave Plasma. Diamond Relat. Mater. 2011, 20, 232−237. (8) Sankaran, K. J.; Kumar, N.; Kurian, J.; Ramadoss, R.; Chen, H. C.; Tyagi, A. K.; Dash, S.; Lee, C. Y.; Tai, N. H.; Lin, I. N. Improvement in Tribological Properties by Modification of Grain Boundary and Microstructure of Ultrananocrystalline Diamond Films. ACS Appl. Mater. Interfaces 2013, 5, 3614−3624. (9) Zhu, W.; Kochanski, G. P.; Jin, S. Low-Field Electron Emission from Undoped Nanostructured Diamond. Science 1998, 282, 1471− 1473. (10) Cheng, H. F.; Horng, C. C.; Chiang, H. Y.; Chen, H. C.; Lin, I. N. Modification on the Microstructure of Ultrananocrystalline Diamond Films for Enhancing Their Electron Field Emission Properties via a Two-Step Microwave Plasma Enhanced Chemical Vapor Deposition Process. J. Phys. Chem. C 2011, 115, 13894−13900. (11) Teng, K. Y.; Huang, P. C.; Shih, W. C.; Lin, I. N. The Induction of a Graphitic-like Phase by Fe-Coating/Post-Annealing Process To Improve the Electron Field Emission Properties of Ultrananocrystalline Diamond Films. Diamond Relat. Mater. 2012, 24, 188−194. (12) Lin, I. N.; Chen, Y. H.; Cheng, H. F. Modification of Emission Properties of Diamond Films Due to Surface Treatment Process. Diamond Relat. Mater. 2000, 9, 1574−1581. (13) Nebel, C. E. Surface-Conducting Diamond. Science 2007, 318, 1391−1392. (14) Seshan, V.; Ullien, D.; Castellanos-Gomez, A.; Sachdeva, S.; Murthy, D. H. K.; Savenije, T. J.; Ahmad, H. A.; Nunney, T. S.; Janssens, S. D.; Haenen, K.; Nesládek, M.; van der Zant, H. S. J.; Sudhölter, E. J. R.; de Smet, L. C. P. M. Hydrogen Termination of CVD Diamond Films by High-Temperature Annealing at Atmospheric Pressure. J. Chem. Phys. 2013, 138, 234707. (15) Wang, M.; Simon, N.; Decorse-Pascanut, C.; Bouttemy, M.; Etcheberry, A.; Li, M. S.; Boukherroub, R.; Szunerits, S. Comparison of the Chemical Composition of Boron-Doped Diamond Surfaces upon Different Oxidation Processes. Electrochim. Acta 2009, 54, 5818−5824. (16) Ristein, J. Surface Science of Diamond: Familiar and Amazing. Surf. Sci. 2006, 600, 3677−3689. (17) Ariano, P.; Baldelli, P.; Carbone, E.; Gilardino, A.; Lo Guidice, A.; Lovisolo, D.; Manfredotti, C.; Novara, M.; Sternschulte, H.; Vittone, E. Cellular Adhesion and Neuronal Excitability on Functionalised Diamond Surfaces. Diamond Relat. Mater. 2005, 14, 669−674. (18) Tachiki, M.; Kaibara, Y.; Sumikawa, Y.; Shigeno, M.; Banno, T.; Song, K. S.; Umezawa, H.; Kawarada, H. Diamond Nanofabrication and Characterization of Biosensing Applications. Phys. Status Solidi A 2003, 199, 39−43. (19) Nebel, C. E.; Kato, H.; Rezek, B.; Shin, D.; Takeuchi, D.; Watanabe, H.; Yamamoto, T. Electrochemical Properties of Undoped Hydrogen Terminated CVD Diamond. Diamond. Relat. Mater. 2006, 15, 264−268. (20) Lafosse, A.; Berlin, M.; Michaelson, S.; Azria, R.; Akhvlediani, R.; Hoffmann, A. Surface Defects Induced by In-Situ Annealing of Hydrogenated Polycrystalline Diamond Studied by High Resolution Electron Energy Loss Spectroscopy. Diamond Relat. Mater. 2008, 17, 949−953. (21) Kawarada, H. Hydrogen-Terminated Diamond Surface and Interfaces. Surf. Sci. Rep. 1996, 26, 205−259. (22) Hoffmann, R.; Kriele, A.; Obloh, H.; Hees, J.; Wolfer, M.; Smirnov, W.; Yang, N.; Nebel, C. E. Electrochemical Hydrogen

■

CONCLUSIONS The effects of H2 gas and H2 plasma treatments on the structural and EFE properties of UNCD films were studied. The UNCD films were grown from an Ar/CH4 gas mixture by an MPECVD system. To study the effects of bonding characteristics on the EFE properties, two different treatment methods were used. The first method was the H2 plasma treatment, which was carried out in an MPECVD system at 600 °C. The second method was the H2 gas treatment, which was done in the nonplasma quartz tube reactor connected with a H2 gas supply line at 600 °C. We have observed that H2-gas-treated UNCD films possess superior EFE properties compared to the H2-plasma-treated UNCD films. TEM investigations revealed that the induction of a-C (or graphite) grain boundary phases in the whole region of the UNCD films due to H2 gas treatment created conduction channels for the electrons to transport from the bottom to the top of the films and emit to vacuum very easily, resulting in the enhancement of EFE and PI properties. On the contrary, H2 plasma treatment induced graphitic grain boundary phases only in the surface region of the UNCD films. The interior region is still mostly sp3diamond phases with lower amounts of a-C (or graphite) grain boundary phases, ensuing in inferior EFE and PI properties. As a consequence, the induction of a-C (or graphite) grain boundaries in the whole region of UNCD films due to H2 gas treatment is the genuine factor for the enhancement of the EFE and the PI properties. The results support the potential use of UNCD films in a broad spectrum of applications in flat panel and microplasma display technologies.

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The thank the National Science Council for financial support through Project Nos. NSC 101-2221-E-007-064-MY3 and NSC 102-2112-M-032-006.

■

REFERENCES

(1) Zuiker, C.; Krauss, A. R.; Gruen, D. M.; Pan, X.; Li, J. C.; Csencsits, R.; Erdemir, A.; Bindal, C.; Fenske, G. Physical and Tribological Properties of Diamond Films Grown in Argoncarbon Plasmas. Thin Solid Films 1995, 270, 154−159. (2) Hupert, M.; Muck, A.; Wang, J.; Stotter, J.; Cvackova, Z.; Haymond, S.; Show, Y.; Swain, G. M. Conductive Diamond ThinFilms in Electrochemistry. Diamond Relat. Mater. 2003, 12, 1940− 1949. (3) Zhou, D.; Krauss, A. R.; Qin, L. C.; McCauley, T. G.; Gruen, D. M.; Corrigan, T. D.; Chang, R. P. H.; Gnaser, H. Synthesis and Electron Field Emission of Nanocrystalline Diamond Thin Films Grown from N2/CH4 Microwave Plasmas. J. Appl. Phys. 1997, 82, 4546−4550. (4) Krauss, A. R.; Auciello, O.; Ding, M. Q.; Gruen, D. M.; Huang, Y.; Zhirnov, V. V.; Givargizov, E. I.; Breskin, A.; Chechen, R.; Shefer, E.; Konov, V.; Pimenov, S.; Karabutov, A.; Rakhimov, A.; Suetin, N. 14550

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551

ACS Applied Materials & Interfaces

Research Article

Termination of Boron-Doped Diamond. Appl. Phys. Lett. 2010, 97, 052103. (23) Manfredotti, Ch.; Bonino, P.; Pierre, M. D. L.; Vittone, E.; Manfredotti, C. About Orientation Dependence of Physico-Chemical Properties of HPHT Diamond Surface Thermally Treated in H2 and D2 Environments. Diamond Relat. Mater. 2010, 19, 279−283. (24) Ando, T.; Ishii, M.; Kamo, M.; Sato, Y. Thermal Hydrogenation of Diamond Surfaces Studied by Diffuse Reflectance FourierTransform Infrared, Temperature-Programmed Desorption and Laser Raman Spectroscopy. J. Chem. Soc., Faraday Trans. 1993, 89, 1783−1789. (25) Fizzotti, F.; Lo Giudice, A.; Manfredotti, C.; Manfredotti, C.; Castellino, M.; Vittone, E. Diamond Surface Conductivity after Exposure to Molecular Hydrogen. Diamond Relat. Mater. 2007, 16, 836−839. (26) Dong, C. L.; Chen, S. S.; Chiou, J. W.; Chen, Y. Y.; Guo, J. H.; Cheng, H. F.; Lin, I. N.; Chang, C. L. Effect of Surface Treatments on the Electronic Properties of Ultra-Nanocrystalline Diamond Films. Diamond Relat. Mater. 2008, 17, 1150−1153. (27) Lee, Y. C.; Pradhan, D.; Lin, S. J.; Chia, C. T.; Cheng, H. F.; Lin, I. N. Effect of Surface Treatment on the Electron Field Emission Property of Nano-Diamond Films. Diamond Relat. Mater. 2005, 14, 2055−2058. (28) Panda, K.; Sankaran, K. J.; Panigrahi, B. K.; Tai, N. H.; Lin, I. N. Direct Observation and Mechanism for Enhanced Electron Emission in Hydrogen Plasma-Treated Diamond Nanowire Films. ACS Appl. Mater. Interfaces 2014, 6, 8531−8541. (29) Fowler, R. H.; Nordheim, L. Electron Emission in Intense Electric Fields. Proc. R. Soc. London, Ser. A 1928, 119, 173−181. (30) Ferrari, A. C.; Robertson, J. Origin of the 1150 cm−1 Raman Mode in Nanocrystalline Diamond. Phys. Rev. B 2001, 63, 121405. (31) Corrigan, T. D.; Gruen, D. M.; Krauss, A. R.; Zapol, P.; Chang, R. P. H. The Effect of Nitrogen Addition to Ar/CH4 Plasmas on the Growth, Morphology and Field Emission of Ultrananocrystalline Diamond. Diamond Relat. Mater. 2002, 11, 43−48. (32) Ferrari, A. C.; Robertson, J. Interpretation of Raman Spectra of Disordered and Amorphous Carbon. Phys. Rev. B 2000, 61, 14095− 14107. (33) Kalish, R. Doping of Diamond. Carbon 1999, 37, 781−785. (34) Heng, H.; Jun, H. X.; Wen, B. B.; Hu, C. X. Effects of Annealing Time on the Microstructural and Electrochemical Properties of BDoped Nanocrystalline Diamond Films. Acta Phys. Sin. 2012, 61, 148101. (35) Tsai, W. C.; Wang, S. J.; Chang, C. L.; Chen, C. H.; Ko, R. M.; Liou, B. W. Improvement of Field Emission Characteristics of Tungsten Oxide Nanowires by Hydrogen Plasma Treatment. EPL 2008, 84, No. 16001. (36) You, J. B.; Zhang, X. W.; Cai, P. F.; Dong, J. J.; Gao, Y.; Yin, Z. G.; Chen, N. F.; Wang, R. Z.; Yan, H. Enhancement of Field Emission of the ZnO Film by the Reduced Work Function and the Increased Conductivity via Hydrogen Plasma Treatment. Appl. Phys. Lett. 2009, 94, 262105. (37) Zheng, Z.; Liao, L.; Yan, B.; Zhang, J. X.; Gong, H.; Shen, Z. X.; Yu, T. Enhanced Field Emission from Argon Plasma-Treated UltraSharp α-Fe2O3 Nanoflakes. Nanoscale Res. Lett. 2009, 4, 1115−1119. (38) Robertson, J. Mechanism of Electron Field Emission from Diamond and Diamond-Like Carbon. IVMC 98, Asheville, NC, July 19−24, 1998; IEEE: Piscataway, NJ, 1998; pp 162−163. (39) Lieberman, M. A.; Lichtenberg, A. J. Principles of Plasma Discharges and Materials Processing, 2nd ed.; John Wiley & Sons: Hoboken, NJ, 2005. (40) Chapman, B. Sputtering and Plasma Etching; John Wiley & Sons: New York, 1980. (41) Srinivasu, K.; Sankaran, K. J.; Leou, K. C.; Tai, N. H.; Lin, I. N. Microplasma Device Architectures with Various Diamond Nanostructures and Their Characteristics. Plasma Sources Sci. Technol. 2014, submitted for publication.

(42) Villalpando, I.; John, P.; Porro, S.; Wilson, J. I. B. Hydrogen Plasma Etching of Diamond Films Deposited on Graphite. Diamond Relat. Mater. 2011, 20, 711−716. (43) Joseph, P. T.; Tai, N. H.; Lee, C. Y.; Niu, H.; Pong, W. F.; Lin, I. N. Field Emission Enhancement in Nitrogen-Ion-Implanted Ultrananocrystalline Diamond Films. J. Appl. Phys. 2008, 103, 043720. (44) Gruen, D. M.; Liu, S.; Krauss, A. R.; Luo, J.; Pan, X. Fullerenes as Precursors for Diamond Film Growth without Hydrogen or Oxygen Additions. Appl. Phys. Lett. 1994, 64 (12), 1502−1504. (45) Kovarik, P.; Bourdon, E. B. D.; Prince, R. H. Electron-EnergyLoss Characterization of Laser-Deposited a-C, a-C:H, and Diamond Films. Phys. Rev. B 1993, 48, 12123−12129. (46) Prawer, S.; Peng, J. L.; Orwa, J. O.; McCallum, J. C.; Jamieson, D. N.; Bursill, L. A. Size Dependence of Structural Stability in Nanocrystalline Diamond. Phys. Rev. B 2000, 62, R16360−R16363. (47) Sankaran, K. J.; Kurian, J.; Chen, H. C.; Dong, C. L.; Lee, C. Y.; Tai, N. H.; Lin, I. N. Origin of a Needle-like Granular Structure for Ultrananocrystalline Diamond Films Grown in a N2/CH4 plasma. J. Phys. D: Appl. Phys. 2012, 45, 365303. (48) Teng, K. Y.; Chen, H. C.; Chiang, H. Y.; Horng, C. C.; Cheng, H. F.; Sankaran, K. J.; Tai, N. H.; Lee, C. Y.; Lin, I. N. The Role of Nano-Graphite Phase on the Enhancement of Electron Field Emission Properties of Ultrananocrystalline Diamond Films. Diamond Relat. Mater. 2012, 24, 126−133. (49) Yamaguchi, H.; Masuzawa, T.; Nozue, S.; Kudo, Y.; Saito, I.; Koe, J.; Kudo, M.; Yamada, T.; Takakuwa, Y.; Okano, K. Electron Emission from Conduction Band of Diamond with Negative Electron Affinity. Phys. Rev. B 2009, 80, 165321. (50) Geis, M. W.; Deneault, S.; Krohn, K. E.; Marchant, M.; Lyszczarz, T. M.; Cooke, D. L. Field Emission at 10 Vcm−1 with Surface Emission Cathodes on Negative-Electron-Affinity Insulators. Appl. Phys. Lett. 2005, 87, 192115. (51) Cui, J. B.; Stammler, M.; Ristein, J.; Ley, L. Role of Hydrogen on Field Emission from Chemical Vapor Deposited Diamond and Nanocrystalline Diamond Powder. J. Appl. Phys. 2000, 88, 3667−3673.

14551

dx.doi.org/10.1021/am503823n | ACS Appl. Mater. Interfaces 2014, 6, 14543−14551